Introduction
The management of patients who are receiving a direct oral anticoagulant (DOAC) and
require an elective surgery or invasive procedure is an increasingly common and challenging
clinical scenario.[1 ] The clinical scope of this problem is increasing, as DOACs are replacing warfarin
as the recommended first-line anticoagulant treatment option for stroke prevention
in patients with atrial fibrillation.[2 ]
[3 ] Moreover, the number of patients with atrial fibrillation is increasing due to an
aging population, and such older patients are more likely to require a surgery/procedure
than younger patients.[4 ] It is estimated that 12 to 15% of patients with atrial fibrillation will require
anticoagulant interruption for an elective surgery/procedure annually.[5 ]
[6 ] The perioperative management of such patients is pertinent to many clinicians, including
internists, surgeons, anaesthetists, family physicians, and dentists.[4 ]
[7 ]
Managing such patients is challenging because evidence-based guidelines to inform
best practices are lacking. Although the American and European Societies of Regional
Anesthesia, the American College of Cardiology and the American Heart Association
have provided guidelines for the perioperative management of DOAC-treated patients,
these documents can only be considered expert-opinion guidance, as they are not anchored
on findings from prospective studies that assess perioperative management strategies.[8 ]
[9 ]
[10 ] Only one prospective study, to our knowledge, has assessed a standardized perioperative
management approach in patients who required an elective surgery/procedure, but this
study was limited as it included only patients on dabigatran.[11 ]
Given the magnitude of the clinical problem coupled with the paucity of practice-informing
prospective clinical studies, the Perioperative Anticoagulant Use for Surgery Evaluation
(PAUSE) study (NCT02228798) was designed. The primary study aim is to establish a
safe, standardized protocol for the perioperative management of patients with atrial
fibrillation who are receiving a DOAC (dabigatran, rivaroxaban or apixaban) and require
treatment interruption for an elective surgery/procedure. Herein, we describe the
rationale for the PAUSE study design and analysis plan, and elaborate on the reasoning
behind the standardized pre-procedure DOAC interruption and post-procedure DOAC resumption
protocols.
Study Methods
Study Aims, Design and Rationale
The primary aim of PAUSE is to demonstrate the safety of a standardized but patient-specific
protocol for the perioperative management of DOACs. This protocol is adjusted according
to DOAC type, patient's renal function and surgery/procedure-related bleeding risk
to optimize patient safety, and does not involve heparin bridging anticoagulation.[1 ] The PAUSE study requires approval by institutional review boards of all participating
clinical sites and written informed consent from all participating patients.
We hypothesize that the perioperative management protocol is ‘safe for patient care’
in patients with atrial fibrillation, defined by an expected low risk for major bleeding
(i.e. 1%; 80% power to exclude 2%) and an expected low risk for arterial thromboembolism
(i.e. 0.5%; 80% power to exclude 1.5%). These expected risks were based on the available
literature at the time of the protocol development (in 2013), which showed perioperative
risks of major bleeding of 1 to 3% and risks of arterial thromboembolism of 0.4 to
1.5% in warfarin- and DOAC-treated patients who, like patients in PAUSE, would have
anticoagulant interruption for an elective surgery/procedure and did not receive heparin
bridging.[12 ]
[13 ]
The secondary study aim is to determine the effect of the DOAC interruption protocol
on the level of residual anticoagulation when measured by DOAC-specific coagulation
tests, comprising the dilute thrombin time (dTT) for dabigatran and anti-factor Xa
assays for rivaroxaban and apixaban.[14 ] For all DOACs, the effect of DOAC interruption on non-specific coagulation tests,
comprising the prothrombin time (PT), the activated partial thromboplastin time (aPTT)
and thrombin time (TT), is also assessed.[15 ]
[16 ] In an exploratory manner, we postulated that the pre-procedure DOAC interruption
protocol is associated with a high proportion (90%) of patients with a minimal residual
DOAC-associated anticoagulant effect just before a surgery/procedure, which we define
as a dTT or anti-factor Xa level <50 ng/mL.
The PAUSE study uses a prospective cohort design, with three parallel groups for dabigatran,
rivaroxaban and apixaban (edoxaban and betrixaban were not available for clinical
use when PAUSE was designed), to assess a standardized but patient-specific perioperative
management protocol. A non-randomized, prospective cohort design was chosen based
on the following considerations: First, a cohort design is appropriate to assess the
safety of a clinical management strategy that is expected to be associated with low
risks of adverse clinical events, and when there is sufficient statistical power to
exclude clinically important higher risk of events.[17 ]
[18 ] Although cohort studies have the potential for patient selection bias, for example
if presumed high-risk patients are more likely to be excluded, such bias was deemed
unlikely in PAUSE because in a precursor perioperative dabigatran study, 90% of eligible
patients consented to participate.[11 ] Moreover, the lack of a standard-of-care approach for the perioperative management
of DOAC-treated patients would make it unlikely that patient groups would be systematically
excluded, especially when the aim of PAUSE is to establish best practices. Second,
a randomized trial design was not considered because a true control group would be
lacking, as there is no established or de facto standard of care for perioperative DOAC management. For example, a comparator group
wherein DOACs are interrupted 5 days pre-procedure in all patients, as is done with
warfarin, would be a second experimental arm. Third, a cluster randomized trial design,
whereby clinical sites would allocate patients to the PAUSE protocol or usual care,
was not considered because usual care would frequently be the same or similar to the
PAUSE protocol in some clinical sites, thereby precluding a distinction between the
two perioperative management strategies.
Patient Population and Rationale
All of the following inclusion criteria have to be satisfied for patient study eligibility:
(1) age ≥18 years; (2) receiving dabigatran (150 or 110 mg twice daily), rivaroxaban
(20 or 15 mg daily) or apixaban (5 or 2.5 mg twice daily) for stroke prevention in
atrial fibrillation; (3) undergoing an elective surgery or procedure that requires
temporary interruption of anticoagulant therapy and (4) ability to adhere to the pre-procedure
DOAC interruption protocol. Patients are excluded if they have ≥1 of the following
criteria: (1) creatinine clearance (CrCl) <30 mL/min for dabigatran- and rivaroxaban-treated
patients, or CrCl <25 mL/min for apixaban-treated patients, as estimated by the Cockcroft-Gault
formula,[19 ] as these DOACs are contraindicated if the CrCl is below these cut-points; (2) cognitive
impairment or psychiatric illness that might preclude reliable post-procedure follow-up
and documentation of outcome events; (3) inability or unwillingness to provide informed
consent and (4) previous participation in the PAUSE study. In PAUSE, we use the Cockcroft-Gault
formula to estimate patient eligibility according to renal function because this formula
has been used to determine patient eligibility for the randomized trials that assessed
DOACs.[20 ]
[21 ]
[22 ]
[23 ] In everyday practice, it is also recommended to use the Cockcroft-Gault formula
to estimate renal function and, in turn, assess patient eligibility for DOACs, and
to determine the dose regimen for rivaroxaban.[24 ]
The PAUSE study focuses on patients with atrial fibrillation because this is the dominant
clinical indication for long-term DOAC therapy, and such patients comprise 70 to 80%
of patients who are assessed for perioperative anticoagulant management.[25 ] Patients with a mechanical heart valve are excluded because DOACs are contraindicated
in such patients, but patients with bioprosthetic heart valves or non-rheumatic valvular
heart disease are not excluded as DOACs can be used in such patients.[2 ]
[3 ] We do not include patients with venous thromboembolism as the sole indication for
DOAC therapy because they constitute a separate population who tend to be younger
and with fewer comorbidities than patients with atrial fibrillation, and are more
likely to develop venous rather than arterial thromboembolic adverse outcomes.[26 ]
[27 ] Moreover, inclusion of such patients would, in effect, create a dichotomous patient
population that would affect the generalizability of the study results.
Classification of Bleeding Risk Associated with Surgery/Procedure and Thromboembolism
Risk
The classification of surgery/procedure types as low bleeding risk or high bleeding
risk is an essential part of the pre-procedure assessment, as this determines the
timing of DOAC interruption and resumption. Procedures classified as high bleeding
risk have a longer interruption and resumption interval (2 days) compared with those
having a low bleeding risk (1 day). As there is no established classification scheme
to distinguish surgery/procedure-associated bleeding risk, the approach used in PAUSE
is empiric but based on the bleeding risk classification used in the BRIDGE trial.[4 ] In addition, the PAUSE protocol allows flexibility in the procedure-related bleeding
risk classification to account for real-life situations. For example, a patient having
surveillance colonoscopy would have low-bleeding risk management pre-procedure (1
day off DOAC), but if multiple or large polyps are removed during the procedure, this
patient would be allowed to have high-bleeding risk management where DOAC resumption
is delayed for 48 to 72 hours.[28 ]
[29 ]
The PAUSE study documents patients' CHADS2 and CHA2 DS2 -VASc scores,[30 ]
[31 ] but these scores do not affect perioperative anticoagulant management. Thus, no
patient receives therapeutic-dose heparin bridging in PAUSE, irrespective of the perceived
thromboembolic risk. The no-bridging approach adopted in PAUSE is supported by the
recent publication of the BRIDGE trial in warfarin-treated patients,[4 ] and related studies in DOAC- and warfarin-treated patients,[7 ]
[32 ] showing that a no-bridging strategy is non-inferior to bridging to prevent thromboembolism
and superior to a bridging strategy for the prevention of major bleeding.
Peri-procedural Management Protocol and Rationale
The overall aim is to develop a standardized but patient-specific management protocol,
which accounts for the DOAC taken, patient's renal function, and surgery/procedure
type and which is easy to implement in clinical practice. The design of the pre-procedure
DOAC interruption protocol, as shown in [Fig. 1 ], was the most challenging aspect of the protocol development. Particular attention
is given to patients having a high bleeding risk surgery/procedure or neuraxial anaesthesia,
so that the DOAC interruption interval results in minimal to no residual anticoagulant
effect at the time of the surgery/anaesthesia.
Fig. 1 PAUSE protocol for DOAC interruption before a surgery/procedure. Solid arrows indicate
usual DOAC dose taken on those days; shaded area indicates no DOAC taken on those
days. DOAC, direct oral anticoagulant; PAUSE, the Perioperative Anticoagulant Use
for Surgery Evaluation.
Pre-procedure DOAC Interruption Protocol
Three sources of information were considered in developing the pre-procedure DOAC
interruption interval: (1) the elimination half-life of each DOAC, as reported for
dose regimens used for stroke prevention in atrial fibrillation;[33 ]
[34 ]
[35 ]
[36 ] (2) the DOAC manufacturers' recommended pre-procedure interruption intervals and
(3) the available literature relating to perioperative management of DOAC-treated
patients.
Another issue was whether the pre-procedure DOAC interruption would be expressed as
an hour-based (i.e. 24 or 48–72 hours) or day-based interval (1 or 2–3 days) between
the last drug dose and the surgery/procedure. We chose the ‘day-based interval’, comprising
1 day off for a low bleeding risk and 2 days off for a high bleeding risk surgery/procedure
because it offered a simple protocol that would be easy to apply in everyday practice
and easy to follow by patients. In addition, it allowed a slightly longer DOAC interruption
interval than the ‘hour-based interval’, which would be particularly important to
ensure minimal to no residual anticoagulant effect in patients having a high bleeding
risk surgery/procedure.[37 ]
To illustrate the advantage of this approach used in PAUSE, consider a patient with
atrial fibrillation (CrCl, 60 mL/min) who is having elective hip replacement on a
Monday at 8 am and is taking dabigatran, 150 mg twice daily, at 9 am and 6 pm. If the ‘hour-based interval’ is used, the last dabigatran dose would be Friday 6
pm , to allow a 48-hour interval between the last dose and the surgery. If, on the other
hand, the surgery was on a Monday at noon, the last dabigatran dose would be Saturday
at 9 am to have a 48-hour interruption interval. By comparison, the ‘day-based interval’
would simply mean the last dabigatran dose would be on Friday (2 days off before the
surgery) irrespective of the timing of surgery on Monday. If the surgery was on Monday
at noon, the PAUSE approach would also allow a longer dabigatran interruption interval
of 66 hours (with last dose on Friday 6 pm ), instead of a 51-hour interval (with last dose on Saturday 9 am ). Given the dabigatran elimination half-life of 12 to 14 hours, the 2-day off approach
would correspond to approximately five elimination half-lives between the last dose
and the time of surgery and would provide reassurance to the surgeon and anesthetist
of minimal to no remaining anticoagulant effect.
In patients having a low bleeding risk surgery/procedure, a 1-day off (or 36–42 hours)
interval between the last DOAC dose and the surgery/procedure would be acceptable,
as there would be a minimal residual anticoagulant at the time of the surgery/procedure
effect corresponding to three to four elimination half-lives between the last DOAC
dose and the surgery/procedure. In all patients, no DOAC would be taken on the day
of the surgery/procedure.
Patient's renal function was also considered in determining the pre-procedure DOAC
interruption interval. If a patient was taking dabigatran, which has 75 to 80% renal
clearance, and had a CrCl 30 to 50 mL/min, the duration of interruption was extended
to 2 days in patients having a low bleeding risk procedure and to 4 days in patients
having a high bleeding risk procedure. If a patient was taking rivaroxaban or apixaban
and had a CrCl 30 to 50 mL/min, there was no extension in the interruption interval,
as these DOACs have only 25 to 33% renal clearance.
Post-procedure DOAC Resumption Protocol
As shown in [Fig. 2 ], DOACs would be resumed 1 day (∼24 hours) after a low bleeding risk surgery/procedure
and 2 to 3 days (∼48–72 hours) after a high bleeding risk surgery/procedure. This
flexible, patient-specific approach reflects the variable bleeding risk associated
with different surgery/procedure types and is designed to minimize the risk for post-procedure
bleeding. Previous studies suggested a high post-procedural bleeding risk if therapeutic-dose
heparins are uniformly re-initiated approximately 24 hours after a procedure irrespective
of procedural bleeding risk.[38 ]
Fig. 2 PAUSE protocol for DOAC resumption after a surgery/procedure. † First post-procedure dose delayed at least 24 hours post-procedure; ‡ first post-procedure dose delayed at least 48 hours post-procedure. Solid arrows indicate
usual DOAC dose taken on those days; shaded area indicates no DOAC taken on those
days; shaded portion of arrow indicates optional resumption of DOAC on day +2 or +3.
DOAC, direct oral anticoagulant; PAUSE, the Perioperative Anticoagulant Use for Surgery
Evaluation.
For patients in whom DOAC resumption was delayed for 2 to 3 days post-procedure and
those who were considered high risk for venous thromboembolism, the PAUSE protocol
permits the use of postoperative thromboprophylaxis with low-dose unfractionated or
low-molecular-weight heparin. Similarly, a heparin could be used in patients who cannot
take medications by mouth, for example in the setting of a post-operative ileus.
Clinical Outcomes and Rationale
The primary clinical outcomes are arterial thromboembolism, comprising stroke (ischemic
or haemorrhagic), systemic embolism or transient ischemic attack and major bleeding;
the secondary outcomes comprise all-cause death, acute coronary syndrome, venous thromboembolism,
clinically relevant non-major bleeding and minor bleeding. The period of observation
begins from the time of DOAC interruption until 30 days post-procedure. All clinical
outcomes are defined based on objective diagnostic criteria (see Appendix A ), and are adjudicated by an independent events adjudication committee, which is blinded
to the DOAC patients are receiving.
The rationale for the primary outcomes was to identify thromboembolic events related
to atrial fibrillation and anticoagulant interruption and to identify bleeding outcomes
that would necessitate either prolonged anticoagulant interruption or re-operation.
The classification of non-major bleeds into clinically relevant non-major or minor
reflected the variable severity of non-major bleeds, for example some requiring medical
attention or intervention. We acknowledge that distinguishing between clinically relevant
non-major and minor bleeds might pose challenges given that some bleeding is expected
in a perioperative setting, especially after major surgery.[39 ]
Laboratory Outcomes and Rationale
The primary laboratory outcomes vary according to the DOAC assessed. For dabigatran-treated
patients, this is the dTT, as it is considered the most reliable and precise coagulation
test to measure the anticoagulant effect of dabigatran.[14 ] For rivaroxaban- and apixaban-treated patients, it is DOAC-calibrated anti-factor
Xa levels.[14 ] The cut-point (50 ng/mL) used to define a clinically acceptable low residual anticoagulant
effect was decided upon by consensus among investigators, as there is no widely accepted
cut-point that defines a safe level of residual anticoagulation to allow a surgery/procedure
to proceed. Alternative cut-points, such as assay-specific lower limits of detection
of anticoagulant effect, will also be considered.
The pre-procedure blood samples used in PAUSE are processed at participating clinical
sites, using standardized methods, and shipped to the Special Coagulation Laboratory
at McMaster University Medical Center for centralized and standardized measurement
of the PT, aPTT, TT, dTT and anti-factor Xa levels. The blood processing methods,
the coagulation tests used and reference values are described in Appendix B .
Study Hypothesis and Sample Size Determination
The primary study hypothesis is that the PAUSE protocol is safe for the perioperative
management of each DOAC; that is the risk for major bleeding at a patient level is
1% and the risk for arterial thromboembolism is 0.5%. The sample size calculation
is based on an estimated risk of major bleeding at 1%, with a non-inferiority proportion
of 2% (i.e., the largest proportion that is still considered non-inferior to the reference
value of 1%). The required sample size is 987 patients per DOAC, which provides 80%
power at the 5% significance level (one sided). The total number of required patients
per DOAC was increased by 10% to 1,097 to anticipate patients with a cancelled surgery/procedure
and those lost to follow-up. We are more confident about estimates of major bleeding
than arterial thromboembolism and, consequently, major bleeding is the primary determinant
of sample size, but the sample chosen is such that the expected range of arterial
thromboembolism risks is also addressed.
Statistical Analysis Plan
We shall use descriptive statistics to report on baseline patient characteristics
as follows: continuous variables will be reported as median (and interquartile range
[IQR]) or mean (and standard deviation), and categorical variables will be reported
as frequency or proportions (and 95% CIs). The proportions, and associated 95% CIs,
for major bleeding and arterial thromboembolism will be calculated at the patient
level for each DOAC. Non-inferiority tests for one proportion will be performed to
examine if, at the patient level, the proportion of major bleeding per DOAC is <2%
and the proportion of arterial thromboembolism per DOAC is <1.5%. Additional secondary
analyses are planned to assess outcomes according to different DOAC doses, comprising
dabigatran 150 and 110 mg, rivaroxaban 20 and 15 mg, and apixaban 5 and 2.5 mg.
For the laboratory outcomes, we will determine in an exploratory manner the median
(IQR) of non-specific coagulation test values (PT, aPTT, TT) for all DOACs. We will
determine median (IQR) dTT values for patients taking dabigatran and median (IQR)
anti-factor Xa levels for patients taking rivaroxaban or apixaban. We will also determine
the proportion of patients with dTT and anti-factor Xa levels below certain cut-points,
including <50 ng/mL.
Discussion
The PAUSE trial is a prospective cohort study that aims to determine if a standardized
but patient-specific perioperative management protocol is safe for patients with atrial
fibrillation on DOACs who need anticoagulant interruption for an elective surgery/procedure.
Safety is defined, for each DOAC, by the management protocol being associated with
perioperative risks of major bleeding of 1.0% (80% power to exclude 2%) and risks
of arterial thromboembolism of 0.5% (80% power to exclude 1.5%). These risk estimates
are similar to those observed in studies involving DOAC- and warfarin-treated patients
who required perioperative anticoagulant interruption and did not receive heparin
bridging.[7 ]
[32 ]
[40 ]
[41 ]
From a study design standpoint, although randomized trials remain the methodological
reference standard to investigate pharmacologic or other management interventions,
the lack of a plausible comparator group in the domain of perioperative DOAC management
precluded adopting such a study design. The prospective cohort design used in PAUSE
is also appropriate when expected risks of clinical events are low. Moreover, demonstrating
a small treatment effect with different perioperative management strategies when risks
of clinical events are expected to be low is unlikely to be clinically meaningful,
for example with a 33% reduction in either major bleeding (from 1.5 to 1.0%) or arterial
thromboembolism (from 1.0 to 0.67%).
The PAUSE study should be considered within the context of related studies assessing
the perioperative management of DOAC-treated patients. As shown in [Table 1 ], there are four sub-studies from the randomized trials that assessed DOACs for stroke
prevention in atrial fibrillation in which there was an analysis of perioperative
outcomes in patients who required an elective surgery/procedure during the course
of these trials.[5 ]
[40 ]
[41 ] However, these were retrospective analyses that did not include standardized perioperative
management protocols, with the exception of the RELY study in which a pre-procedure
interruption protocol was introduced about half-way through the trial. The prospective
Dresden Registry assessed 30-day post-procedure outcomes in 595 rivaroxaban-treated
patients who had 863 surgery/procedures.[42 ] A limitation of this study was that perioperative DOAC management was not standardized
and only 10% of patients underwent a major procedure. Finally, the prospective dabigatran
cohort study is the only study that incorporates a standardized pre- and post-procedure
management protocol.[11 ] However, this study was limited to patients on dabigatran and only 96 (of 541) patients
had a major surgery, in whom more robust data on perioperative management is needed.
Finally, none of these studies routinely measured a residual anticoagulant effect
at the time of the surgery/procedure. Another study assessed a residual anticoagulant
effect after the pre-procedure interruption of dabigatran or rivaroxaban in 65 patients,
but the interruption interval, which was not standardized, varied from 1 to 168 hours.[43 ]
Table 1
Clinical and study characteristics, and perioperative outcomes from sub-studies or
primary studies involving DOAC-treated patients with atrial fibrillation who had an
elective surgery/procedure
Study and patient characteristics and outcomes assessed
Study (reference)
RE-LY sub-study[5 ]
[7 ]
ROCKET AF sub-study[41 ]
ARISTOTLE sub-study[40 ]
ENGAGE AF-TIMI sub-study[44 ]
Dresden study[42 ]
Canadian study[11 ]
PAUSE study
Study design
Retrospective cohort
Retrospective cohort
Retrospective cohort
Retrospective cohort
Patient registry
Prospective cohort
Prospective cohort
DOAC studied
Dabigatran
Rivaroxaban
Apixaban
Edoxaban
Rivaroxaban[a ]
Dabigatran
Dabigatran, rivaroxaban, apixaban
Number of patients studied with perioperative management
2,691
2,165
1,960
4,825
595
541
3,300
Number of surgeries or procedures studied
2,691
3,393
2,877
4,825
863
541
3,300
Percent of patients having major/high bleeding risk surgeries or procedures
17%
18%
10%
23%
10%
40%
33%
Mean age, y
72.7
73.0
71.0
71.6
74.0
72.2
TBD
Mean CHADS2 score
2.1
3.4
2.1
2.8
n/a
n/a
TBD
Standardized perioperative management protocol
No[b ]
No
No
No
No
Yes
Yes
Perioperative use of heparin bridging allowed
Yes
Yes
Yes
Yes
Yes
No
No
Pre-procedure measurement of residual anticoagulant effect
No
No
No
No
No
No[c ]
Yes
Perioperative incidence of stroke or systemic embolism (%)[d ]
0.30
0.35
0.35
0.66
1.0[e ]
0.20
TBD
Perioperative incidence of major bleeding (%)[d ]
[f ]
2.6
1.0
1.6
1.9
1.2
1.8
TBD
Abbreviations: CrCl, creatinine clearance; CHADS2 , c ongestive heart failure, h ypertension, a ge ≥75 years, d iabetes, prior s troke or transient ischemic attack; DOAC, direct oral anticoagulant; n/a, not available;
TBD, to be determined.
a 17% of patients receiving rivaroxaban for venous thromboembolism.
b Suggested pre-procedure management protocol introduced part-way through the RE-LY
trial.
c In a sub-study, 181 patients had pre-procedure measurement of residual anticoagulant
effect.[45 ]
d Period of observation varies across studies but mainly 30-days post-procedure.
e Comprising adverse cardiovascular outcomes.
f Major bleeding definition not standardized across studies.
There are two ongoing multi-centre patient registries of DOAC-treated patients who
are having an elective surgery/procedure: the Observatory of Invasive Procedures and
Bleeding in Patients Treated with New Oral Anticoagulants (GIHP-NACO, NCT02185027)
study in France and the Periprocedural Direct Oral Anticoagulant Management (RA-ACOD,
NCT03182218) study in Spain. Taken together, these patient registries will provide
helpful data regarding perioperative outcomes in DOAC-treated patients, but additional
prospective cohort studies with standardized anticoagulant management such as PAUSE
are needed. To our knowledge, the only other prospective perioperative DOAC management
studies (not registered in ClinicalTrials.gov) involve patients with atrial fibrillation
who receive only apixaban and patients with prior venous thromboembolism who receive
any DOAC. Finally, there are ongoing studies assessing DOAC continuation in patients
who are having cardiac pacemaker or implantable cardiac defibrillator procedures,
which is an emerging management option with such procedures, but involves patients
who differ from those eligible for PAUSE.
There are potential limitations to the PAUSE study design. First, it may not be possible
to enrol the same targeted number of patients (1,092) who are receiving dabigatran,
rivaroxaban or apixaban. In clinical practice, the global use of each DOAC will vary
over time depending on factors that include drug costs, new clinical trial data and
drug marketing. The strategy adopted in PAUSE is to continue the study until the targeted
sample size is attained for each DOAC so that the safety of the PAUSE protocol can
be ascertained for each DOAC with the same statistical power. Second, it may be argued
that the interruption interval and decision to proceed to a surgery/procedure should
be driven by the residual anticoagulant effect, as measured by DOAC-specific coagulation
tests.[46 ] In PAUSE, the interruption interval is driven by the pharmacokinetic properties
of each DOAC. This approach has been used for the perioperative management of patients
who are receiving warfarin or other vitamin K antagonists, as well as unfractionated
or low-molecular-weight heparins, in which laboratory tests at the time of a surgery/procedure
are not mandated to ensure minimal or no residual anticoagulant effect.[6 ]
[13 ] Moreover, requiring laboratory testing prior to a surgery/procedure is problematic
because DOAC-specific assays are costly, add complexity to pre-surgical assessments,
are not routinely available and, even if available, there is uncertainty regarding
which assay should be used and what would be considered a safe anticoagulant level
cut-point (e.g. <50 or <30 ng/mL).[47 ] Nevertheless, this may be an acceptance problem for interventional specialties such
as surgery and anaesthesia that routinely perform coagulation testing prior to procedures.
Third, there may be under-representation of patients at high risk for thromboembolism
or those having high bleeding risk surgery or neuraxial anaesthesia. We believe this
is unlikely because PAUSE involves a simple, patient-friendly protocol, and we anticipate
that a high proportion of eligible patients will be recruited, with adequate representation
of various patient and surgery risk categories.
In summary, the PAUSE study will address a common and important unmet clinical need
and has the potential to establish a standard-of-care approach for the perioperative
management of DOAC-treated patients. The PAUSE management protocol was designed to
be easily applied in everyday practice as it is standardized, allowing application
across institutions, and also flexible to the real-world eventualities that typify
the perioperative clinical setting.
